To meet the requirements for faster performance, the characteristic dimensions of features of integrated circuit devices have continued to be decreased. Manufacturing of devices with smaller feature sizes introduces new challenges in many of the processes conventionally used in semiconductor fabrication. One of the most important of these fabrication processes is photolithography.
Organic polymer films, particularly those that absorb at the i-line (365 nm) and g-line (436 nm) wavelengths conventionally used to expose photoresists, and at the recently used 157 nm, 193 nm, 248 nm wavelengths, have been employed or are being tested as anti-reflective coatings. However, the fact that the organic ARC's share many chemical properties with the organic photoresists can limit usable process sequences. Furthermore, ARC's, including both organic and inorganic ARC's, may intermix with photoresist layers. Organic and inorganic ARC's can mix with photoresist layers if they are not sufficiently baked or cured.
One solution to avoid intermixing is to introduce thermosetting binders as additional components of organic ARC's, as described for example in U.S. Pat. No. 5,693,691 to Flaim et al. Dyes may also be incorporated in organic ARC's, as well as, optionally, additional additives such as wetting agents, adhesions promoters, preservatives, and plasticizers, as described in U.S. Pat. No. 4,910,122 to Arnold et al. Another attempt to avoid intermixing is found in U.S. Pat. No. 6,268,108 issued to Iguchi et al. However, the compositions for forming antireflective coatings found in Iguchi must be irradiated with actinic rays in order to produce an acid, which in turn activates a crosslinking reaction. Even though these previous patents may address some of the issues with intermixing, the problem of the lack of 86- to 90-degree uniformity on the resist edges because of the coupled ARC layer has not been addressed in the prior art.
Photoresists and anti-reflective coatings can also influence one another to the extent that the chemical properties of the anti-reflective coating and/or the resist material can lead the resist to “fall over” once a pattern has been developed into the resist. In other words, the patterned resist sidewall can't maintain an approximate 90-degree angle with respect to the anti-reflective coating after photoresist developing. Instead the resist will take on a 120-degree or an 80-degree angle with respect to the anti-reflective coating. These imperfections are also an indication that photoresist materials and anti-reflective coatings are not necessarily chemically, physically or mechanically compatible.
Photoresists and anti-reflective coatings can also have substandard or unacceptable etch selectivity or stripping selectivity. Poor etch selectivity and/or stripping selectivity can lead to low etch rates for the film. Poor etch selectivity can also lead to poor transfer of critical dimensions from the printing step(s) through the etch step(s). Attempts have been made at improving the etch rate by providing highly absorbing substances with subsititution groups that can condense the silane compound to specific silane compounds, as seen in JP Patent Application No.: 2001-92122 published on Apr. 6, 2001. However, the etch selectivity obtained with these reactive compounds are not sufficient for most photoresists and anti-reflective coatings and require additional chemical reaction steps that may not be necessary.
In addition, photoresists and anti-reflective coatings often have difficulty with fill bias and voiding in via structures to the point where any planarization of the surface is severely compromised. Oftentimes, the two goals of increasing etch selectivity and minimizing fill bias and voiding directly conflict with one another, which is why it's important to review and understand the goals of each group of applications. Also, to sufficiently fill and planarize via arrays requires that a relatively thick anti-reflective coating exist. If the ARC coating is organic, such a thick coating will further compromise the accurate transfer of the as patterned critical dimension through the film stack.
Via first trench last (VFTL) copper dual damascene patterning through a low dielectric constant (less than about 3) material or ultra low dielectric constant (less than about 2) material can be very difficult. One of the problems with this type of patterning is the selective removal of the sacrificial fill material from the low dielectric constant materials. Previous work has shown that Si—O fill materials (either UV absorbing or transparent) are the optimum materials platform, if the dielectric layer is Si—O based.
In order to improve the removal selectivity of the sacrificial fill material it can be chemically weakened relative to the dielectric material. A porogen or a high boiling solvent can be added to the fill material to weaken it; however, in order to achieve photoresist developer resistance the Si—O based fill material either needs to be baked to or at a sufficiently high temperature to ensure crosslinking or the porogen content needs to be lowered. Both of these methods designed to achieve photoresist developer resistance work with respect to strengthening the fill material, but the removal selectivity of the fill material is significantly decreased.
A class of materials that can be used as an anti-reflective layer is spin-on-glass (SOG) compositions containing a dye. Yau et al., U.S. Pat. No. 4,587,138, disclose a dye such as basic yellow #11 mixed with a spin-on-glass in an amount approximately 1% by weight. Allman et al. U.S. Pat. No. 5,100,503 disclose a cross-linked polyorganosiloxane containing an inorganic dye such as TiO2, Cr2O7, MoO4, MnO4, or ScO4, and an adhesion promoter. Allman additionally teaches that the spin-on-glass compositions also serve as a planarizing layer. However, the spin-on-glass, dye combinations that have been disclosed to date are not optimal for exposure to the deep ultraviolet, particularly 248 and 193 nm, light sources that are coming into use to produce devices with small feature sizes. Furthermore, not all dyes can be readily incorporated into an arbitrary spin-on-glass composition. Also, even though these ARC's are chemically different than the previously mentioned organic ARC's, the coupled resist layers can still suffer from “falling over” after being developed, as based on the chemical, physical, and mechanical incompatibility of the ARC layer and the resist layer—which is a common problem when trying to couple resist materials and anti-reflective coatings.
In developing anti-reflective coatings that can a) absorb strongly and uniformly in the ultraviolet spectral region; b) keep the resist material from “falling over” and expanding outside or contracting inside of the intended resist line and c) be impervious to photoresist developers and methods of production of spin-on glass anti-reflective coatings, Baldwin et al developed several anti-reflective coatings that are superior to conventional anti-reflective coatings, including those materials and coatings found in US Issued U.S. Pat. Nos. 6,268,457 issued on Jul. 31, 2001; 6,365,765 issued on Apr. 2, 2002; 6,368,400 issued on Apr. 9, 2002; U.S. patent application Ser. Nos. 09/491,166 filed Jan. 26, 2000; 10/012,651 filed Nov. 5, 2001; 10/012,649 filed Nov. 5, 2001; 10/001,143 filed Nov. 15, 2001; PCT Applications Serial Nos: PCT/US00/15772 filed on Jun. 8, 2000; WO 02/06402 filed on Jul. 12, 2001; PCT/US01/45306 filed on Nov. 15, 2001; Pending PCT Application Serial No.: PCT/US02/35101 filed on Oct. 31, 2002; European Patent Applications Serial No. 00941275.0 filed on Jun. 6, 2000; and 01958953.0 filed on Jul. 17, 2001, which are all commonly assigned and incorporated herein by reference in their entirety. However, with all of these materials, it would be beneficial to be able to modify the materials, coatings and films described therein to improve etch selectivity and/or stripping selectivity, improve the lithography performance and to minimize fill bias.
Therefore, an absorbing/anti-reflective coating and lithography material that a) absorbs strongly and uniformly in the ultraviolet spectral region, b) can keep the resist material from “falling over” and expanding outside or contracting inside of the intended resist line, c) would be impervious to photoresist developers and methods of production of the SOG anti-reflective coating described; d) can satisfy any goals of increasing etch selectivity and/or stripping selectivity and e) can satisfy any goals of minimizing fill bias and voiding in via structures; f) can form solutions that are stable and have a good shelf life; g) is compatible to various lithographic patterning techniques, including those that utilize ArF; h) can be applied to a surface by any suitable application method, such as spin-on coating or chemical vapor deposition (CVD); i) is capable of via fill and planarization; j) has good wet etch and dry etch rates; and k) can be utilized in a number of applications, components and materials, including logic applications and flash applications.